Geospatial analysis of groundwater quality in the major cities of Pakistan

The aim of this study is to assess the spatial variability of groundwater quality parameters. Geographic information systems multivariate statistical techniques and nonparametric kriging methods were used to analyze the correlational structure and spatial pattern of groundwater quality parameters in the major cities of Pakistan. The hydro-chemical results of 366 water samples were taken from the Pakistan Council of Research in Water Resources (PCRWR) report 2015-2016 of 25 major cities of Pakistan. The correlation matrix was used to identify the highly correlated groundwater quality parameters. The principal component analysis and cluster analysis categorized the quality parameters according to the variation. The results indicated that seven water quality parameters including electric conductivity, calcium (Ca), magnesium (Mg), hardness, sodium (Na), sulfate (SO4) and total dissolved solids were found exceeding the permissible limits of World Health Organization (WHO). Due to highly skewed data, nonparametric kriging methods were used to estimate the probability of concentration of groundwater quality parameters and to produce prediction maps. Cross-validation statistics demonstrated that the indicator kriging method showed better performance than ordinary kriging and co-kriging methods for mapping groundwater quality parameters. Overall water quality results showed that only 113 (31%) out of 366 water samples were suitable for drinking, whereas 253 (69%) were not safe drinking water. This study highlights the use of nonparametric kriging methods for non-normal data and concludes that solid waste and sewage systems should be developed to reduce the contamination of groundwater resources.